Plots corresponding to Moving-Average(weekly), Moving-Average(Monthly), Moving-Average(Quarterly)               

Calculating Daily returns based on percent                                                 
	Percent change can provide insights into the volatility and distribution of        
	the stock returns over time                                                        

pip install pandas_ta

   List of Financial Indicators employed:-  

['RSI: Relative Strength Index',
 'ADL: Accumulation/Distribution Line',
 'ATR: Average True Range',
 'MOM: Momentum',
 'MFI: Money Flow Index',
 'ROC: Rate of Change',
 'OBV: On-Balance Volume',
 'CCI: Commodity Channel Index']

Text(0.5, 0, 'Correlation')

A very brief Explanation

• RSI (Relative Strength Index) - measures the momentum of a stock's price and helps to determine overbought or oversold conditions.
  Parameters required to determine: close
  

• ADL (Accumulation/Distribution Line) - measures the flow of money into or out of a stock.
  Parameters required to determine: high,low,close,volume

• ATR (Average True Range) - shows the volatility of a stock and helps to identify potential price breakouts or breakdowns.
  Parameters required to determine: high, low, close

• MOM (Momentum) - measures the rate of change in a stock's price over a specified period of time.
  Parameters required to determine: close

• MFI (Money Flow Index) - combines volume and price data to measure buying and selling pressure.
  Parameters required to determine: high, low, close, volume

• ROC (Rate of Change) - calculates the percentage change in a stock's price over a specified period of time.
  Parameters required to determine: close

• OBV (On-Balance Volume) - measures buying and selling pressure by adding volume on up days and subtracting volume on down days.
  Parameters required to determine: Close, volume

• CCI (Commodity Channel Index) - measures a stock's deviation from its average price over a specified period of time, helping to identify potential trend reversals.
  Parameters required to determine: high, low, close

(2601, 10)

(2601, 5)

 The parameter Adjusted Close i.e. Adj Close will be the target veriable for model building and prediction              
	as it is the closing price after adjustments for all applicable splits and dividend distributions.  

     The data for futher processing is of shape (2601, 5)     

Preparing data for modelling

Two distinct methodologies have been employed,

1. Univariate LSTM modeling --> Only one feature is used to predict future value of stock.
2. Multivariate LSTM modeling --> Several features will be considered to predict future value of stock.

	The Train-Validation-Test Split         

--------------------********--------------------
Length of train data is 2081,
Length of Validation data is 312,
Length of Test data is 208
--------------------^^^^^^^^--------------------

scaler_ys = MinMaxScaler(feature_range=(0,1)) scaler_y = scaler_ys.fit(np.array(train_.iloc[:training_len,0]).reshape(-1,1))

(2081,)

((2081, 5), (2601, 5), (2601, 5))

Training data is dated from 2013-01-21 to 2021-07-01 i.e. 2081 days
---------------------------------------------------------------------------
Validation will begin from 2021-07-02 to 2022-10-03 i.e 312 days
---------------------------------------------------------------------------
lastly, Test data is from 2022-10-04 to 2023-08-04 i.e 208 days
--------------------------------^^^^^^^^^^^--------------------------------

LSTM uses previous day's data to predict future data points hence we generate columns of n previous dates

Data Preprocessing for LSTM modelling

For Multivariate LSTM modelling

The splitting is now shaped as follows after running it throught the generate_data function
	The data is currently structured as a 3D tensor with a dimension of 10 timestamps and 5 features
----------------------------------------------------
train shape: (2070, 10, 5) (2070,) (2070,)
validation: (301, 10, 5) (301,) (301,)
test shape: (197, 10, 5) (197,) (197,)

For Univariate LSTM modelling

The splitting is now shaped as follows after running it throught the generate_data_univerate function
	10 columns generated, having adj close price prior to 10 days will be used as X_train while next-day will be y_train

The corresponding are the X,Y and Date array shapes
----------------------------------------------------
train shape: (2070, 10) (2070,) (2070,)
validation: (301, 10) (301,) (301,)
test shape: (197, 10) (197,) (197,)

 Converting data into a 3d array for Univariate LSTM 

train data shape for univerate: X: (2070, 10, 1) | y: (2070,)
validation shape for univerate: X:  (301, 10, 1) | y: (301,)
test data shape for univerate:  X:  (197, 10, 1) | y: (197,)

Model: "univariate_LSTM"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 lstm_13 (LSTM)              (None, 10, 128)           66560     
                                                                 
 lstm_14 (LSTM)              (None, 64)                49408     
                                                                 
 dropout_11 (Dropout)        (None, 64)                0         
                                                                 
 dense_13 (Dense)            (None, 25)                1625      
                                                                 
 dense_14 (Dense)            (None, 1)                 26        
                                                                 
=================================================================
Total params: 117,619
Trainable params: 117,619
Non-trainable params: 0
_________________________________________________________________

Model: "Multivariate-LSTM"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 lstm_62 (LSTM)              (None, 10, 128)           68608     
                                                                 
 dropout_50 (Dropout)        (None, 10, 128)           0         
                                                                 
 lstm_63 (LSTM)              (None, 64)                49408     
                                                                 
 dropout_51 (Dropout)        (None, 64)                0         
                                                                 
 dense_75 (Dense)            (None, 25)                1625      
                                                                 
 batch_normalization_13 (Bat  (None, 25)               100       
 chNormalization)                                                
                                                                 
 dense_76 (Dense)            (None, 12)                312       
                                                                 
 dense_77 (Dense)            (None, 1)                 13        
                                                                 
=================================================================
Total params: 120,066
Trainable params: 120,016
Non-trainable params: 50
_________________________________________________________________

deleted model
deleted history

[datetime.date(2010, 1, 4), '---', datetime.date(2023, 8, 4), 2666]

(2666, 7, 5)

Conforming the data to the specified model through fitting

1. For Univariate-LSTM
2. For Multivariate-LSTM

Epoch 1/15
27/27 [==============================] - 1s 23ms/step - loss: 0.0041 - val_loss: 0.0049
Epoch 2/15
27/27 [==============================] - 1s 22ms/step - loss: 0.0027 - val_loss: 0.0015
Epoch 3/15
27/27 [==============================] - 1s 24ms/step - loss: 0.0033 - val_loss: 0.0018
Epoch 4/15
27/27 [==============================] - 1s 21ms/step - loss: 0.0020 - val_loss: 0.0013
Epoch 5/15
27/27 [==============================] - 1s 21ms/step - loss: 0.0023 - val_loss: 0.0014
Epoch 6/15
27/27 [==============================] - 1s 21ms/step - loss: 0.0015 - val_loss: 0.0012
Epoch 7/15
27/27 [==============================] - 1s 21ms/step - loss: 0.0017 - val_loss: 0.0011
Epoch 8/15
27/27 [==============================] - 1s 21ms/step - loss: 0.0011 - val_loss: 0.0013
Epoch 9/15
27/27 [==============================] - 1s 21ms/step - loss: 0.0014 - val_loss: 0.0011
Epoch 10/15
27/27 [==============================] - 1s 21ms/step - loss: 9.1076e-04 - val_loss: 0.0011
Epoch 11/15
27/27 [==============================] - 1s 21ms/step - loss: 0.0010 - val_loss: 0.0010
Epoch 12/15
27/27 [==============================] - 1s 23ms/step - loss: 8.6512e-04 - val_loss: 0.0011
Epoch 13/15
27/27 [==============================] - 1s 23ms/step - loss: 9.4404e-04 - val_loss: 0.0011
Epoch 14/15
27/27 [==============================] - 1s 22ms/step - loss: 8.7484e-04 - val_loss: 0.0011
Epoch 15/15
27/27 [==============================] - 1s 22ms/step - loss: 9.2262e-04 - val_loss: 0.0010

                              Univariate LSTM Model                              
7/7 [==============================] - 0s 7ms/step
65/65 [==============================] - 1s 8ms/step

Considering 10 previous date data for model training the performance is as follows
After 40 epochs,
Train RMSE = 12.168
Test RMSE = 14.963

Best Test fit was obtained after

Traning the Multivariate model for:-

1. 50 epoch with batch size 32.....40ep, 17e
2. 50 epoch with batch size 64
3. 50 epoch with batch size 128
4. 50 epoch with batch size 256
5. 10 epoch with batch size 384
6. 10 epoch with batch size 768

Gave the best fit on test data with RMSE as under

                              Multivariate LSTM Model                              
7/7 [==============================] - 0s 7ms/step
65/65 [==============================] - 0s 7ms/step

Considering 10 previous date data for model training the performance is as follows
After 230 epochs,
Train RMSE = 11.432
Test RMSE = 11.042

Note on model performance

It is not unusual for the validation loss and training loss to differ in LSTM modeling.
The most important thing to look for is how well the model is performing on the test data.
If the test data fit and test RMSE are good, that indicates that the model is able to accurately predict new data.

Scaled_Train RMSE = 0.020 Scaled_Test RMSE = 0.023 Train RMSE = 10.783 Test RMSE = 12.069

The following algorithm predicts the anticipated price fluctuations of a stock by implementing a preset date as input.

This algorithm calculates the stock's performance from the succeeding "duration_".
Upon executing the algorithm,
the user will be prompted to enter the desired "Date" and the number of days for predictions.
Subsequently, a graph representing the generated predictions will be displayed.
Additionally, there is an option to save the predictions as a CSV file for further reference.

Funtions used to create the above algorithm are as follows :

1. predict_condition -- Checks if the dataframe has dates succeding to the give date
2. pred -- Generates the model input data based on the provided date
3. predict_for -- Predicts the price for n days based on user input
4. plot_fututre -- Plots graphs based on the predictions(in red) in continuation to previous data points(in blue)

Univariate model prediction

• Adjusted Close which is also a target veriable

                              Forecasting with Multivariate LSTM Model                              

The Model predicts based on the data of 10 
kindly,
wait for predictions


future value of stock on 2023-08-05 is predicted to be 629.108